#!/usr/bin/env python
# coding: utf-8

# In[1]:


import os
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
from IPython.core.display import display, HTML
get_ipython().run_line_magic('matplotlib', 'inline')


# In[2]:


import folium
from folium.plugins import FastMarkerCluster


# In[3]:


pd.options.display.max_columns = None
# display(HTML("<style>.container { width:100% !important; }</style>"))


# In[4]:


CSV_PATH = os.path.join('data', 'hacknight_ticket_sample_data_2015.csv')
df = pd.read_csv(CSV_PATH,low_memory=False, parse_dates=['issue_date', 'ticket_queue_date'])


# In[5]:


CSV_PATH = os.path.join('data', 'hacknight_sample_data_geocode.csv')
addrs_df = pd.read_csv(CSV_PATH)
addrs_df['address'] = addrs_df['address'] + ', chicago, il'


# In[6]:


geocoded_df = pd.merge(left=df, right=addrs_df, how='inner', on='address')
geocoded_df.head()


# The map below shows where the most prolific ticketer (Officer #728) wrote their tickets. As you zoom in on the map, the map will automatically show finer detail as to where the tickets were written.

# In[7]:


# Which CPD officers wrote at least 50 tickets in 2015
prolific_CPD = geocoded_df.loc[geocoded_df['unit_description'] == 'CPD', 'officer'].value_counts()\
                [geocoded_df.loc[geocoded_df['unit_description'] == 'CPD','officer'].value_counts() >= 50]


# In[8]:


chi_lat = 41.8
chi_lon = -87.75


# In[9]:


chi_map = folium.Map(location=[chi_lat, chi_lon], zoom_start=10)

# Officer #728 is the officer that wrote the most tickets in 2015
subset_df = geocoded_df[geocoded_df['officer'].isin(prolific_CPD.index.tolist())]

chi_map.add_child(FastMarkerCluster(subset_df[['lat', 'lng']].values.tolist()))
chi_map


# In[10]:


# What violations occurred at least 100 times?
query = geocoded_df['violation_description'].value_counts()
query[(query > 100)]


# **Where were people ticketed for not having a parking permit?**

# In[11]:


chi_map = folium.Map(location=[chi_lat, chi_lon], zoom_start=10)

# Where were people ticketed for not having a parking permit?
subset_df = geocoded_df[geocoded_df['violation_description'].isin(['EXPIRED PLATES OR TEMPORARY REGISTRATION'])]

chi_map.add_child(FastMarkerCluster(subset_df[['lat', 'lng']].values.tolist()))
chi_map


# In[12]:


chi_map = folium.Map(location=[chi_lat, chi_lon], zoom_start=10)

# Where were people ticketed for not having a parking permit?
subset_df = geocoded_df[geocoded_df['violation_description'].isin(['EXPIRED PLATES OR TEMPORARY REGISTRATION'])]

chi_map.add_child(FastMarkerCluster(subset_df[['lat', 'lng']].values.tolist()))
chi_map


# In[13]:


# What are the rare cares that were ticketed? (>=5 to eliminate unknown makers, <=50 to get rare cars)
query = geocoded_df['vehicle_make'].value_counts()
query[(query >= 5) & (query <= 50)]


# In[14]:


# What are the rare cares that were ticketed? (>=5 to eliminate randos, <=50 to get rare cars)
rare_cars = query[(query >= 5) & (query <= 50)].index.tolist()


# **Where were rare cars (as defined above) ticketed?**

# In[15]:


chi_map = folium.Map(location=[chi_lat, chi_lon], zoom_start=10)

# Where were rare cars (as defined above) ticketed?
subset_df = geocoded_df[geocoded_df['vehicle_make'].isin(rare_cars)]

chi_map.add_child(FastMarkerCluster(subset_df[['lat', 'lng']].values.tolist()))
chi_map


# **Where were Ferraris and Maseratis ticketed?**

# In[16]:


chi_map = folium.Map(location=[chi_lat, chi_lon], zoom_start=10)

# Where were Ferraris and Maseratis ticketed?
subset_df = geocoded_df[geocoded_df['vehicle_make'].isin(['FERR', 'MASE'])]

chi_map.add_child(FastMarkerCluster(subset_df[['lat', 'lng']].values.tolist()))
chi_map


# In[17]:


geocoded_df['notice_level'].value_counts()


# **Where were vehicles seized?**

# In[18]:


chi_map = folium.Map(location=[chi_lat, chi_lon], zoom_start=10)

# Where were vehicles seized?
subset_df = geocoded_df[geocoded_df['notice_level'] == 'SEIZ']

chi_map.add_child(FastMarkerCluster(subset_df[['lat', 'lng']].values.tolist()))
chi_map


# In[19]:


query = geocoded_df['license_plate_type'].value_counts()
query[query >= 20]


# **Where were vehicles with temporary license plates ticketed?**

# In[20]:


chi_map = folium.Map(location=[chi_lat, chi_lon], zoom_start=10)

# Where were vehicles with temporary license plates ticketed?
subset_df = geocoded_df[geocoded_df['license_plate_type'] == 'TMP']

chi_map.add_child(FastMarkerCluster(subset_df[['lat', 'lng']].values.tolist()))
chi_map


# In[21]:


chi_map = folium.Map(location=[chi_lat, chi_lon], zoom_start=10)

# Officer #728 is the officer that wrote the most tickets in 2015
subset_df = geocoded_df[geocoded_df['officer'] == '653']

chi_map.add_child(FastMarkerCluster(subset_df[['lat', 'lng']].values.tolist()))
chi_map


# In[22]:


chi_map = folium.Map(location=[chi_lat, chi_lon], zoom_start=10)

# Officer #728 is the officer that wrote the most tickets in 2015
subset_df = geocoded_df[geocoded_df['officer'] == '790']

chi_map.add_child(FastMarkerCluster(subset_df[['lat', 'lng']].values.tolist()))
chi_map


# In[ ]:





# In[ ]: